[MITgcm-support] python analysis tools

[Oliver Jahn]

Hi Ryan,

I have a core module, fa.py, that defines a FacetArray, basically a list
of arrays that behaves like a single array.  It is modelled after the
MaskedArray class in numpy.  The first index is the face index, the
others index each array.  You can slice all faces at once, mostly useful
for dropping some rows/columns, e.g., for derivatives,

  f[...,1:] - f[...,:-1]

There is a companion module, exchange.py, that implements exchanges a la
exch2.  It is initialized from the kind of connectivity data you find in
data.exch2 (slightly modified).  A cubed-sphere topology is already
defined and can easily be modified for llc.

I also attach a basic plotting module that plots multiple facets in a
single quadmesh-like object (so one can easily change color ranges,
...).

With these core modules, a lot of things can be done like in flat
geometries.

Oliver


On 2014-11-08 13:11, Ryan Abernathey wrote:
> Ok, well I was hoping someone would say it was already done. ;)
> 
> I will start working on it, but I suspect it will take a while. Let me
> know (off list) if you want to help.
> 
> -Ryan
> 
> On Fri, Nov 7, 2014 at 3:17 PM, gael forget <gforget at mit.edu
> <mailto:gforget at mit.edu>> wrote:
> 
>     Hi Ryan,
> 
>     while I cannot be of any help with Python coding at this stage, 
>     I just wanted to state that I support your initiative 100%. 
> 
>     Cheers,
>     Gael
> 
>     On Nov 7, 2014, at 2:58 PM, Ryan Abernathey
>     <ryan.abernathey at gmail.com <mailto:ryan.abernathey at gmail.com>> wrote:
> 
>>     Gael's gcmfaces software is obviously an extremely valuable tool
>>     for analyzing the latest generation of MITgcm simulations on
>>     complex grids. It is an awesome contribution that has enabled many
>>     people to do science with ECCO products who otherwise would have
>>     been intimidated by the technical challenges.
>>
>>     Without entering into a debate on the relative merits of matlab
>>     vs. python, I wanted to ask whether anyone is working on a python
>>     package with capabilities similar to gcmfaces. (Oliver?)
>>
>>     I have written some generic analysis tools for mitgcm simulations
>>     in python
>>     https://bitbucket.org/ryanaberanthey/mitgcmdata
>>     but they can't deal with lat-lon-cap or other complex grids. This
>>     is a project I would consider tackling over the next year, but not
>>     if it is already underway elsewhere.
>>
>>     Thanks,
>>     Ryan
>>     _______________________________________________
>>     MITgcm-support mailing list
>>     MITgcm-support at mitgcm.org <mailto:MITgcm-support at mitgcm.org>
>>     http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> 
>     _______________________________________________
>     MITgcm-support mailing list
>     MITgcm-support at mitgcm.org <mailto:MITgcm-support at mitgcm.org>
>     http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> 
> 
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fa.py
Type: text/x-python
Size: 42915 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20141109/f9df8f72/attachment-0003.py>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: exchange.py
Type: text/x-python
Size: 21964 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20141109/f9df8f72/attachment-0004.py>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: faplot.py
Type: text/x-python
Size: 10182 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20141109/f9df8f72/attachment-0005.py>